Flag job

Report

Freelance Researcher – AI & Machine Learning (Adversarial Financial Fraud Detection)

Location

Chennai, Tamil Nadu, India

JobType

part-time

About the job

Info This job is sourced from a job board

About the role

Workfoster

Website: workfoster.com
Job details:

Job Title: Freelance Researcher – AI & Machine Learning (Adversarial Financial Fraud Detection)


Job Type: Freelancer


Experience Level: Mid–Senior (2–4+ Years)


Domain: AI, Machine Learning, Financial Fraud Detection, Adversarial ML


Role Overview:

We are seeking a highly skilled Freelance AI/ML Researcher to lead advanced research in adversarial financial fraud detection. The role involves designing, replicating, and extending state-of-the-art Transformer-based fraud detection models under data poisoning and adversarial attack scenarios, with a strong focus on robustness, reinforcement learning–based defenses, and reproducible research outcomes.


This engagement is research-intensive and aimed at delivering Ph.D.-level outputs, including publication-ready results and a Scopus-aligned dissertation or research paper.


Key Responsibilities:

  • Lead end-to-end research, design, and implementation of AI/ML-based financial fraud detection systems under adversarial data poisoning conditions.
  • Replicate, benchmark, and extend state-of-the-art Transformer-based fraud detection models using real-world financial transaction datasets.
  • Design and simulate controlled data poisoning attacks, including:

Label flipping

Feature manipulation

Subpopulation poisoning

Backdoor attacks

  • Develop a robust Transformer-based fraud detection framework resistant to adversarial manipulation.
  • Integrate a Reinforcement Learning (RL)–based adaptive defense layer to counter evolving threats, concept drift, and poisoning patterns.
  • Build an ensemble defense architecture combining:

Transformer predictions

RL agent decisions

Confidence scores from state-of-the-art models

  • Conduct comprehensive experimental evaluations under clean and poisoned data conditions using standard fraud detection metrics.
  • Ensure full reproducibility by delivering clean code, experiment logs, datasets, and documentation.
  • Prepare publication-ready research outputs, including implementation details, evaluations, and technical write-ups.


Required Qualifications:

  • Master’s or Ph.D. in Artificial Intelligence, Machine Learning, Computer Science, Data Science, Cybersecurity, or a related field.
  • Strong academic or applied research background in fraud detection, adversarial machine learning, or cybersecurity analytics.
  • Demonstrated expertise in deep learning architectures, particularly Transformers, and reinforcement learning systems.


Required Experience:

  • 2–4+ years of hands-on experience in applied machine learning, fraud detection, or adversarial ML research.
  • Prior experience working with financial transaction datasets or tabular ML problems.
  • Proven ability to replicate published research models and conduct comparative experimental evaluations.


Technical Skills & Tools/:

  • Programming: Python
  • ML / DL Frameworks: TensorFlow, PyTorch
  • Data Processing: NumPy, Pandas, Scikit-learn, large-scale CSV datasets
  • Visualization & Analysis: Matplotlib, Seaborn
  • Experimentation & Reproducibility: Jupyter Notebook, version-controlled ML pipelines


Core Knowledge & Competencies:

  • Deep understanding of supervised learning, deep learning, and Transformer architectures for tabular data.
  • Strong knowledge of adversarial machine learning threats, data poisoning attacks, and defense strategies.
  • Practical expertise in reinforcement learning, including reward design and adaptive decision-making.
  • Ability to conduct statistically sound evaluations, robustness testing, and sensitivity analysis.
  • Strong analytical mindset, independent research capability, and meticulous documentation skills.
  • Excellent written and verbal communication for collaboration with supervisors, reviewers, and research stakeholders.


Deliverables:

  • Fully implemented and evaluated adversarial-robust fraud detection framework
  • Reproducible source code, datasets, experiment logs, and documentation
  • Scopus-ready research paper / dissertation with experimental validation 


Contact Info:

Phone: +91 95661 33822

Email: hr@workfoster.com

Click on Apply to know more.

Skills

Python
Artificial Intelligence
CSV
data science
deep learning
end-to-end
fraud detection
Jupyter Notebook
machine learning
Matplotlib
NumPy
Pandas
Source Code
TensorFlow
Pytorch